[dlacouture]

Lowell, I don't understand your following statement : "[...] pentax has a compensation factor, for the exposure, when it knows the actual maximum apature."

From what I understand, your tests with your Tamron only show that :
1- the Tamron 28-75 has a slight error in apertures, as in A mode, the actual exposure slightly differs from the computed exposure (say 1/100@f/2.8 doesn't give the same exposure than 1/25 @ f/5.6)
2- stop-down exposures with the same Tamron in non-A show the usual wild behavior...

What you state (the "compensation factor") would be only true if you got two different metering with your Tamron when in A mode and manually set at f/2.8.

And from experience, the metering is always the same between A position and the widest aperture, thus negating the "compensation factor" theory...

[dlacouture]

What is interesting is that the change in exposure compensation when adding the TC is exactly the same as the difference (error) in exposure between F2.8, F4 and F5.5 on manual apature lenses. Since the Sigma TCs only pass lens information streight through, it reports F2.8 for maximum apature even with the TC installed again leading to the possibility that the K10/K20 are calibrated with a fudge factor to compensate for the viewing screen.

Does the overexposure occur on the whole aperture range?
I mean : do you have to enter a constant negative Ev comp, regardless of what aperture you'll use?

Because this quite negates the smudge factor again, IMO...

[Lowell Goudge]

Lowell, I don't understand your following statement : "[...] pentax has a compensation factor, for the exposure, when it knows the actual maximum apature."

From what I understand, your tests with your Tamron only show that :
1- the Tamron 28-75 has a slight error in apertures, as in A mode, the actual exposure slightly differs from the computed exposure (say 1/100@f/2.8 doesn't give the same exposure than 1/25 @ f/5.6)
2- stop-down exposures with the same Tamron in non-A show the usual wild behavior...

What you state (the "compensation factor") would be only true if you got two different metering with your Tamron when in A mode and manually set at f/2.8.

this is ecactly the case. metering at F2.8 in A mode is dfifferent than in manual apature mode, but close. The error in metering is the smallest at F2.8 -F4however, if you look at the behavior. therefore considering only the wide open metering is not enough. youo need to consider the difference between wide open metering, and stopped down metering with the same lens (i.e. A and non A modes)

And from experience, the metering is always the same between A position and the widest aperture, thus negating the "compensation factor" theory...

Not true, try the same with an F5.6 lens, and you will see a big change. an F2.8 is too close to correct reading wide open to make the comparison. Edit note, I put my FA 28-105 F4-5.6 on my K10D and in manual mode with the lens wide open set the exposure in A and Manually. In A mode 1/250 at F5.6 in Manual apature 1/100 at F5.6 This is 1 1/3 stops error metering wide open

By changing the focusing screen, and seeing the expsosure change for my 50mm F1.4 this also showed that the linearity of the metering was a function of the focusing screen and apature. What I did not do in all of this is prove the point that changing focusing screen would induce errors in my A lenses by testing exposure on each of those with the *istD screen in the K10D.

[Lowell Goudge]

Does the overexposure occur on the whole aperture range?
I mean : do you have to enter a constant negative Ev comp, regardless of what aperture you'll use?

Because this quite negates the smudge factor again, IMO...

If you are using a TC with an A lens and the lens is in the A position, you are always metering wide open, and calculating the apature for all shots, therefore constant error over the whole apature range using A mode is correct.

I did test the sigma 70-200 in manual apature mode with and without a TC, and the only difference in the behavior was that without the TC the lens metered with one curve, starting at F2.8 close to correct exposure rising to +2 stops over exposed by F8, and by adding the TC wide open the metering started with an identical exposure to the lens on it's own at F5.6 since the 2x TC made the lens an F5.6 this completely fits the theory of the fudge factor in A mode and a non linear performance of metering. Additionally, with the TC and using A mode, the metering has the same error because the camera does not know it is an F5.6 lens at that point.

[Canada_Rockies]

Thanks for all your work, Lowell.

My solution to the problem is to find an aperture that gives me the correct metering for each combination, and memorize it. I then use that and the combinations of equivalent exposure to set the lens up. This could be a real hassle to someone with many lenses. I only have two non-A lenses and two converters, so the combinations are not a long list. I never, ever, use more than one converter on the lens at the same time. Thus, four combinations or me.

In a situation where I can use "sunny 16" I do because it's less of a hassle. A better solution would be to use an external exposure meter.

[Lowell Goudge]

Thanks for all your work, Lowell.

My solution to the problem is to find an aperture that gives me the correct metering for each combination, and memorize it. I then use that and the combinations of equivalent exposure to set the lens up. This could be a real hassle to someone with many lenses. I only have two non-A lenses and two converters, so the combinations are not a long list.

I know what you mean, my list of non A lenses is
24mmTamron F2.5 (although I eally should get the KA adaptor for it)
28mm Vivitar F2.5 (m42)
35mm SMC Takumar F2 (m42)
SMC 50mm F1.4
SMC-M 100mm F4 macro
SMC 105mm F2.8
SMC 135mm F2.5
SMC 300mm F4
Vivitar 400mm F5.6
Vivitar Series 1 70-210 F3.5
Just try to memorize all of those
I'm thinking of making a little wallet sized fold up book, like the one that came with the PZ-1 just to have the curve for each lens, least I forget

I never, ever, use more than one converter on the lens at the same time. Thus, four combinations or me.

In a situation where I can use "sunny 16" I do because it's less of a hassle. A better solution would be to use an external exposure meter.

i've got 2 of those also, but what I have been leaning towards is a katz eye and leaving my *istD which meters perfectly on most of these, as a MF only body

[Canada_Rockies]

As you might notice, only two of my eight lenses have no A position on the aperture ring, and I don't want the camera's metering to go wonky with the six, so using an LL60 or a Katz-Eye is not a solution for me, and it's too cumbersome to meter with the SF-1 or MZ-S and transfer the readings and lens to the k10d. Sigh.

[jeffkrol]

As you might notice, only two of my eight lenses have no A position on the aperture ring, and I don't want the camera's metering to go wonky with the six, so using an LL60 or a Katz-Eye is not a solution for me, and it's too cumbersome to meter with the SF-1 or MZ-S and transfer the readings and lens to the k10d. Sigh.

Just manually meter at a "good" aperature and modify as needed...

Ultimate Exposure Computer

[Canada_Rockies]

Just manually meter at a "good" aperature and modify as needed...

Ultimate Exposure Computer

Earlier in the thread I suggested just that - I agree 100%.

[dlacouture]

Lowell : have you thought about the Matrix vs CW metering? This is what could cause your differences between A position and wide open metering values, as (if I'm not mistaken), going out of the A position forces the metering to CW.

Unfortunately, my k10 went back to pentax for focus tweaking, along with my FA50 and the 18-55, so I cannot test this now...

[Lowell Goudge]

Lowell : have you thought about the Matrix vs CW metering? This is what could cause your differences between A position and wide open metering values, as (if I'm not mistaken), going out of the A position forces the metering to CW.

Unfortunately, my k10 went back to pentax for focus tweaking, along with my FA50 and the 18-55, so I cannot test this now...

not when I'm in spot mode, and / or metering off a uniformly lit surface

I make a specific point of always metering off a uniformly lit uniform grey surface (road or concrete wall) so there is no issue about changes in metering mode interpreting the scene. also as stated above I use spot for most tests.

but you have a really good point for everyone who looks at metering and questions it, matrix metering does all sorts of wonderful things even if yuo only move the camera a few degrees on a complicated lighting scenario. a lot of people question exposure until you point htis out to them. center weighted in many cameras biases the sky to be brighter than the ground, and you need to consider this when you flip the camera on it's side, as to how the metering interprets the sky now that it is not where it should be

[dlacouture]

Damn, you got me here! :-) I'm dry on counter-thoughts...

Okay, I'll have to check on my k10 the day it will come back... Or (maybe ) on my new k20, as the Pentax technical service here in France does propose an upgrade (around 200€).

[stevebrot]

I have been running a series of exposure tests using the stock screen, Katzeye w/o Optibright, and Katzeye with Optibright. I will post the results once I am done.

Steve

[dlacouture]

stevebrot, you're the best!!! I'll be waiting for them!!!

[Lowell Goudge]

stevebrot, you're the best!!! I'll be waiting for them!!!

Me too, as I am sitting on the fence about getting a katzeye and would love to know the answer.